Lamnso', a language with a fairly rigid SVO order, shows asymmetries in the way verb arguments are focused. Subjects are focused post-verbally (in-situ) while non-subjects are focused both pre-verbally (ex-situ) and post-verbally (in-situ). I propose that there are two domains for encoding focus in Lamnso': post-verbal focusing, which will be argued to be an instance of in-situ focus (inside vP), and pre-verbal focusing, which will be argued to be a high focus position (in the CP periphery). The former is used to encode contrastive focus while the latter is used to mark new information focus.